Understanding viewer behaviors in live streaming: the investigations on social live streaming service in China

Abstract

Live streaming is a new form of video-based social media for live sharing. It has become a mass phenomenon in China due to its astonishing popularity and business potential. Live streaming service is special in terms of how the content is created and produced and its monetization model. In a live stream session, there are two levels of social interaction evolving at the same time. Unlike most conventional social media service, there are two parallel of social interaction and the content is co-shaped between the streamer and viewer crowd in a live stream session. With respect to the monetization model, live streaming in China has introduced a novel business model, namely the purchase of virtual gifts which is different from conventional consumption for digital items or services. Hence, understanding viewer behavior of live streaming has both theoretical and practical significance. The insights would contribute to the literature of live streaming and social media usage as well as provide guidance for practitioners to leverage the business potential of live streaming. Although extent research has already started to explore the live streaming phenomenon, currently the research stream is at its infant stage and further systematic investigation is needed. This thesis intends explore the live streaming phenomena by advancing the understanding of viewer behavior in live streaming. Given that the business purposes for such social live streaming service in China is to increase the viewer base and purchase of virtual gifts. This thesis would focus on two viewer behavior namely the viewing behavior and purchase of virtual gifts in live streaming in China. Three specific research questions were proposed: (1) What contextual factors influence viewers to use social live streaming service in general (i.e. viewer stickness and purchase behavior)? (2)Why viewers would stick to a streamer’s live stream session, given the specific cultural context in China? (3) How to make people purchase virtual gifts in live stream session? The first study is the general exploration of what motivate viewers to use live streaming. It started by qualitative exploration to identify the contextual factors(i.e. humor appeal, sex appeal, interactivity and social status display) that may influence viewer stickiness and purchase behavior, and to categorize live streaming types (i.e. event show, knowledge sharing and personal live show). Then the quantitative assessment was followed to examine the effects of contextual factors on the two viewer behaviors in different categories of live streaming. The survey method was deployed and the partial least square structural equation modelling (PLS-SEM) was applied for data analysis. The results suggested that the contextual factors would have various effects on viewer behaviors in different category of live streaming. This study provides the contextual insights on live streaming viewer behavior in general. The second study aims to establish a cultural contextual sensitive framework to explain what increases viewer stickiness to a live stream session. Viewer stickiness reflects the viewers’ sustained interest in a live stream session which is the precondition and foundation of live streaming business potential. Given the social-oriented cultural context of China, the second study first conducted a pre-investigation to confirm the salient effects of viewers’ social perceptions (i.e. perceived proximity to the streamer and sense of belonging to the viewer crowd) on viewers’ immersive and entertaining experience (i.e. flow experience) and viewer stickiness. Based on this, a full model was further developed by incorporating the communication based IT factors (i.e. responsiveness, two-way communication and ) and cultural factor (i.e social orientation) as the explanatory variables for the cultivation of viewers’ social perceptions. The survey method was applied and PLS-SEM was used for data analysis.The results stressed that viewers’social perception can explain the viewer stickiness and flow experience can be transferred from other experience (i.e. social perceptions) in the research context. Study 2 contributes to the knowledge about live streaming watching intention by validating a comprehensive theoretical model to explain viewer stickiness. It also contributes to the knowledge of flow theory by extending the antecedents in social media context. The third study proposed and examined a theoretical framework to explain the purchase of virtual gifts. The purchase of virtual gifts can be perceived to be partially transaction with service provider and partially donation to the relational partner. On the one hand, viewers who enjoy themselves in the live stream session is likely to purchase virtual gifts in exchange of more enjoyable content in a live stream session. Such self-reinforcing motivation was captured by cognitive absorption. On the other hand, viewers purchase virtual gifts in support of the streamer and viewer crowd. The viewers’relational experience and behavioral tendency was captured by para-social interaction with the streamer and viewer crowd participation. The IT-related factors (i.e. viewer-streamer interactivity, viewer-viewer interactivity, streamer-level deep profiling, viewer- level deep profiling and design aesthetics) are included to examine what contributes to the development of viewers’ experience in a live stream session. Survey data was collected and PLS-SEM was applied for data analysis. The results stressed that para-social interaction has no effect on purchase intention, yet cognitive absorption and virtual crowd participation would influence the purchase behavior. This study can be one of the first attempt to explore the mechanism about what makes viewers purchase virtual gifts. It also extends the theories of cognitive absorption and para-social interaction by validating their association. To sum up, the three studies advances the understanding of viewer behavior in live streaming from different angles, and provides practical guidance for live streaming industrial practitioners. Overall, this thesis advances the understanding of viewer behavior in live streaming and contributes to the literature of live streaming or social media in general.

Date of Award	8 Nov 2020
Original language	English
Awarding Institution	Univerisity of Nottingham
Supervisor	Alain Chong (Supervisor), Martin Liu (Supervisor) & Natasa Milic-Frayling (Supervisor)